How to use custom function to add multiple new columns to dataframe?

Question

I have a dataframe and I would like to use a custom function to add multiple new columns to that dataframe. These new columns will be some function of an existing column, but they require the use of a custom function.

I am currently trying to have my custom function return the results in a list, which I then parse into separate columns. This sometimes works by returning a vector of lists, but sometimes this returns a matrix, in which case I get an error like

Error in $<-.data.frame(*tmp*, "z", value = list(1, 2, 2, 3, 3, 4)) : replacement has 2 rows, data has 3

Below is a sample of what I am trying to do.

sample_func <- function(number)
{
list(w = number + 1, u = number + 2)
}

data = data.frame(x = c(1,2,3), y= c(5,6,7))
data$z = sapply(c(1,2,3),sample_func)
data$w = sapply(data$z,"[[","w")
data$u = sapply(data$z,"[[","u")

Sven Hohenstein · Accepted Answer · 2014-01-10 18:43:33Z

2

The function sapply automatically simplifies the result. In this case, you obtain a matrix. You can avoid this behaviour with the argument simplify = FALSE. But it's easier to use lapply because this function doesn't try to simplify the result.

The command

tmp <- lapply(c(1,2,3), sample_func)

returns a list of lists:

[[1]]
[[1]]$w
[1] 2

[[1]]$u
[1] 3


[[2]]
[[2]]$w
[1] 3

[[2]]$u
[1] 4


[[3]]
[[3]]$w
[1] 4

[[3]]$u
[1] 5

You can use the following command to add the new columns to your data frame:

cbind(data, do.call(rbind, tmp))

#   x y w u
# 1 1 5 2 3
# 2 2 6 3 4
# 3 3 7 4 5

Update to address comment:

If possible, you can modify the function to return a data frame.

sample_func <- function(number)
{
  data.frame(w = number + 1, u = number + 2)
}

tmp <- lapply(c(1,2,3), sample_func)

cbind(data, do.call(rbind, tmp))

The result will be a data frame with numeric columns.

edited Jan 10, 2014 at 18:43

answered Jan 10, 2014 at 8:10

Sven Hohenstein

82k17 gold badges150 silver badges173 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

user2374133 Over a year ago

This makes w and u lists, which I then have to convert to numeric. Is there a better way than doing data$w <- as.numeric(data$w)?

Collectives™ on Stack Overflow

How to use custom function to add multiple new columns to dataframe?

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related